Commit Graph

186 Commits

Author SHA1 Message Date
Erik Johnston
9407bcf37a Replace custom DeferredTimeoutError with defer.TimeoutError 2018-09-19 11:07:29 +01:00
Erik Johnston
a334e1cace Update to use new timeout function everywhere.
The existing deferred timeout helper function (and the one into twisted)
suffer from a bug when a deferred's canceller throws an exception, #3842.

The new helper function doesn't suffer from this problem.
2018-09-19 10:39:40 +01:00
Amber Brown
b37c472419
Rename async to async_helpers because async is a keyword on Python 3.7 (#3678) 2018-08-10 23:50:21 +10:00
Matthew Hodgson
5797f5542b WIP to announce deleted devices over federation
Previously we queued up the poke correctly when the device was deleted,
but then the actual EDU wouldn't get sent, as the device was no longer known.
Instead, we now send EDUs for deleted devices too if there's a poke for them.
2018-07-12 01:32:39 +01:00
Amber Brown
49af402019 run isort 2018-07-09 16:09:20 +10:00
Amber Brown
77ac14b960
Pass around the reactor explicitly (#3385) 2018-06-22 09:37:10 +01:00
Amber Brown
071206304d cleanup pep8 errors 2018-05-22 16:54:22 -05:00
Amber Brown
85ba83eb51 fixes 2018-05-22 16:28:23 -05:00
Amber Brown
df9f72d9e5 replacing portions 2018-05-21 19:47:37 -05:00
Richard van der Hoff
6146332387 Merge remote-tracking branch 'origin/develop' into rav/deferred_timeout 2018-04-27 14:18:00 +01:00
Richard van der Hoff
9d2c1b8429 Backport deferred.addTimeout
Twisted 16.0 doesn't have addTimeout, so let's backport it.
2018-04-27 12:52:30 +01:00
Richard van der Hoff
9255a6cb17 Improve exception handling for background processes
There were a bunch of places where we fire off a process to happen in the
background, but don't have any exception handling on it - instead relying on
the unhandled error being logged when the relevent deferred gets
garbage-collected.

This is unsatisfactory for a number of reasons:
 - logging on garbage collection is best-effort and may happen some time after
   the error, if at all
 - it can be hard to figure out where the error actually happened.
 - it is logged as a scary CRITICAL error which (a) I always forget to grep for
   and (b) it's not really CRITICAL if a background process we don't care about
   fails.

So this is an attempt to add exception handling to everything we fire off into
the background.
2018-04-27 11:07:40 +01:00
Richard van der Hoff
1ea904b9f0 Use deferred.addTimeout instead of time_bound_deferred
This doesn't feel like a wheel we need to reinvent.
2018-04-23 00:53:18 +01:00
Adrian Tschira
f63ff73c7f add __bool__ alias to __nonzero__ methods
Signed-off-by: Adrian Tschira <nota@notafile.com>
2018-04-15 20:40:47 +02:00
Richard van der Hoff
d4fb4f7c52 Clear logcontext before starting fed txn queue runner
These processes take a long time compared to the request, so there is lots of
"Entering|Restoring dead context" in the logs. Let's try to shut it up a bit.
2017-11-28 15:26:14 +00:00
Richard van der Hoff
eaaabc6c4f replace 'except:' with 'except Exception:'
what could possibly go wrong
2017-10-23 15:52:32 +01:00
Erik Johnston
ecdd2a3658 Don't start user_directory handling on workers 2017-06-07 12:02:53 +01:00
Erik Johnston
eeb2f9e546 Add user_directory to database 2017-05-31 11:51:01 +01:00
Erik Johnston
62b89daac6 Merge branch 'develop' of github.com:matrix-org/synapse into erikj/repl_tcp_server 2017-04-04 09:46:16 +01:00
Richard van der Hoff
65e1683680 Remove spurious PreserveLoggingContext
In `on_new_room_event`, remove `PreserveLoggingContext` - we can call its
subroutines with the logcontext set.
2017-04-03 15:42:38 +01:00
Richard van der Hoff
feb496056e preserve_fn some deferred-returning things
In `Notifier._on_new_room_event`, `preserve_fn` around its subroutines which
return deferreds, so that it is safe to call it with an active logcontext.
2017-04-03 15:41:17 +01:00
Richard van der Hoff
773e1c6d68 Remove spurious @preserve_fn decorators
Remove `@preserve_fn` decorators on `on_new_room_event`,
`_notify_pending_new_room_events`, `_on_new_room_event`, `on_new_event`, and
`on_new_replication_data` - none of these functions return a deferred, and the
decorator does nothing unless the wrapped function returns a deferred, so the
decorator was a no-op.
2017-04-03 15:14:11 +01:00
Erik Johnston
1df7c28661 Use callbacks to notify tcp replication rather than deferreds 2017-03-31 15:42:51 +01:00
Erik Johnston
7984708a55 Add a simple hook to wait for replication traffic 2017-03-30 11:57:52 +01:00
Erik Johnston
6957bfdca6 Don't recreate so many sets 2017-03-16 11:54:26 +00:00
Erik Johnston
6c82de5100 Format presence events on the edges instead of reformatting them multiple times 2017-03-15 14:27:34 +00:00
Erik Johnston
872e75a3d5 Add some metrics on notifier 2017-03-15 10:56:51 +00:00
Erik Johnston
fd2eef49c8 Reduce spurious calls to generate sync 2017-03-14 16:33:45 +00:00
Erik Johnston
54a79c1d37 Make presence.get_new_events a bit faster
We do this by caching the set of users a user shares rooms with.
2017-02-02 13:07:18 +00:00
Erik Johnston
fbaf868f62 Correctly handle timeout errors 2016-12-09 16:30:29 +00:00
Erik Johnston
be14c24cea Fix rare notifier bug where listeners dont timeout
There was a race condition that caused the notifier to 'miss' the
timeout notification, since there were no other checks for the timeout
this caused listeners to get stuck in a loop until something happened.
2016-12-09 15:43:18 +00:00
Erik Johnston
26072df6af Ensure only main or federation_sender process can send federation traffic 2016-11-23 14:09:47 +00:00
Erik Johnston
f8ee66250a Handle sending events and device messages over federation 2016-11-17 15:48:04 +00:00
Erik Johnston
c0238ecbed Explicitly specify state_key for history_visibility fetching 2016-09-02 14:53:46 +01:00
Erik Johnston
9219139351 Preserve some logcontexts 2016-08-24 11:58:40 +01:00
Erik Johnston
afbf6b33fc defer.returnValue must not be called within Measure 2016-08-19 18:23:44 +01:00
Erik Johnston
39242090e3 Add measure blocks to notifier 2016-08-19 18:23:44 +01:00
Erik Johnston
9da84a9a1e Make AppserviceHandler stream events from database
This is for two reasons:

1. Suppresses duplicates correctly, as the notifier doesn't do any
   duplicate suppression.
2. Makes it easier to connect the AppserviceHandler to the replication
   stream.
2016-08-18 11:54:41 +01:00
Erik Johnston
b9abf3e4e3 Remove dead appservice code 2016-08-17 11:48:23 +01:00
Mark Haines
88625db05f Notify users for events in rooms they join.
Change how the notifier updates the map from room_id to user streams on
receiving a join event. Make it update the map when it notifies for the
join event, rather than using the "user_joined_room" distributor signal
2016-06-07 11:33:36 +01:00
Mark Haines
c626fc576a Move the AS handler out of the Handlers object.
Access it directly from the homeserver itself. It already wasn't
inheriting from BaseHandler storing it on the Handlers object was
already somewhat dubious.
2016-05-31 13:53:48 +01:00
Mark Haines
7e23476814 move filter_events_for_client out of base handler 2016-05-11 13:42:37 +01:00
David Baker
424a7f48f8 Run filter_events_for_client
so we don't accidentally mail out events people shouldn't see
2016-04-27 17:50:49 +01:00
Mark Haines
2a37467fa1 Use google style doc strings.
pycharm supports them so there is no need to use the other format.

Might as well convert the existing strings to reduce the risk of
people accidentally cargo culting the wrong doc string format.
2016-04-01 16:12:07 +01:00
Mark Haines
12904932c4 Hook up adding a pusher to the notifier for replication. 2016-03-15 17:42:03 +00:00
Mark Haines
ddf9e7b302 Hook up the push rules to the notifier 2016-03-03 14:57:45 +00:00
Mark Haines
60a0f81c7a Add a /replication API for extracting the updates that happened on
synapse

This is necessary for replicating the data in synapse to be visible to a
separate service because presence and typing notifications aren't stored
in a database so won't be visible to another process.

This API can be used to either get the raw data by requesting the tables
themselves or to just receive notifications for updates by following the
streams meta-stream.

Returns updates for each table requested a JSON array of arrays with a
row for each row in the table.

Each table is prefixed by a header row with the: name of the table,
current stream_id position for the table, number of rows, number of
columns and the names of the columns.
This is followed by the rows that have been added to the server since
the requester last asked.

The API has a timeout and is hooked up to the notifier so that a slave
can long poll for updates.
2016-03-01 14:49:41 +00:00
Erik Johnston
2c1fbea531 Fix up logcontexts 2016-02-08 14:26:45 +00:00
Daniel Wagner-Hall
d83d004ccd Fix flake8 warnings for new flake8 2016-02-02 17:18:50 +00:00
Erik Johnston
d98a9f2583 Don't use before_token. Its wrong. Use actual limit. 2016-01-31 13:31:15 +00:00
Erik Johnston
13724569ec Deal with None limit 2016-01-29 15:33:44 +00:00
Erik Johnston
4a6eb5eb45 Make /events always return a newer token, if one exists 2016-01-29 15:22:17 +00:00
Erik Johnston
0e07f2e15d Only fetch events for rooms and receipts 2016-01-21 16:10:37 +00:00
Erik Johnston
91a222c66d SYN-606: Peeking does not wake up /events
If a real user attempted to first peek into one room, and then another,
their room event stream would not be woken up for events in the later
room.
2016-01-21 13:22:26 +00:00
Daniel Wagner-Hall
da417aa56d Allow non-guests to peek on rooms using /events 2016-01-20 15:34:07 +00:00
Daniel Wagner-Hall
2ef6de928d Skip, rather than erroring, invalid guest requests
Erroring causes problems when people make illegal requests, because they
don't know what limit parameter they should pass.

This is definitely buggy. It leaks message counts for rooms people don't
have permission to see, via tokens. But apparently we already
consciously decided to allow that as a team, so this preserves that
behaviour.
2016-01-05 18:12:37 +00:00
Mark Haines
f3948e001f Missing yield on guest access auth check
Needs matrix-org/sytest#125 to land first
2015-12-23 14:10:06 +00:00
Daniel Wagner-Hall
468a2ed4ec Return non-room events from guest /events calls 2015-11-12 16:45:28 +00:00
Daniel Wagner-Hall
ca2f90742d Open up /events to anonymous users for room events only
Squash-merge of PR #345 from daniel/anonymousevents
2015-11-05 14:32:26 +00:00
Mark Haines
33b3e04049 Merge branch 'develop' into daniel/removesomelies
Conflicts:
	synapse/notifier.py
2015-11-04 16:01:00 +00:00
Daniel Wagner-Hall
192241cf2a Remove unused arguments and code 2015-11-02 15:10:59 +00:00
Mark Haines
f40b0ed5e1 Inform the client of new room tags using v1 /events 2015-10-29 15:21:09 +00:00
Erik Johnston
f8f3d72e2b Don't make pushers handle presence/typing events 2015-08-24 16:19:43 +01:00
Daniel Wagner-Hall
53d1174aa9 Improve naming 2015-07-20 06:32:12 -07:00
Erik Johnston
bd1236c0ee Consolidate duplicate code in notifier 2015-07-02 11:46:05 +01:00
Erik Johnston
ddf7979531 Add receipts_key to StreamToken 2015-07-02 11:45:44 +01:00
Erik Johnston
2f556e0c55 Fix typo 2015-06-19 16:22:53 +01:00
Erik Johnston
6f6ebd216d PEP8 2015-06-18 17:00:32 +01:00
Erik Johnston
73513ececc Documentation 2015-06-18 16:15:10 +01:00
Erik Johnston
1f24c2e589 Don't bother proxying lookups on _NotificationListener to underlying deferred 2015-06-18 16:09:53 +01:00
Erik Johnston
22049ea700 Refactor the notifier.wait_for_events code to be clearer. Add _NotifierUserStream.new_listener that accpets a token to avoid races. 2015-06-18 15:49:24 +01:00
Erik Johnston
050ebccf30 Fix notifier leak 2015-06-18 11:36:26 +01:00
Mark Haines
554c63ca60 Iterate over the user_streams not the user_ids 2015-05-26 15:03:49 +01:00
Mark Haines
88f1ea36ce Oops, get_rooms_for_user returns a namedtuple, not a room_id 2015-05-21 15:23:58 +01:00
Mark Haines
ad31fa3040 Don't bother sorting by the room_stream_ids, it shouldn't matter which order they are notified in 2015-05-18 14:04:58 +01:00
Mark Haines
1e90715a3d Make sure the notifier stream token goes forward when it is updated. Sort the pending events by the correct room_stream_id 2015-05-18 13:17:36 +01:00
Mark Haines
0ad1c67234 Add some doc-strings to notifier 2015-05-14 14:35:07 +01:00
Mark Haines
084c365c3a Use the current token when timing out a notifier, make sure the user_id is a string in on_new_user_event 2015-05-14 12:03:26 +01:00
Mark Haines
3edd2d5c93 Fix v2 sync, update the last_notified_ms only if there was an active listener 2015-05-14 11:25:30 +01:00
Mark Haines
5e0c533672 Fix metric counter 2015-05-13 17:20:28 +01:00
Mark Haines
f1b83d88a3 Discard unused NotifierUserStreams 2015-05-13 16:54:02 +01:00
Mark Haines
9af432257d Don't set a timer if there's already a result to return 2015-05-13 15:42:13 +01:00
Mark Haines
df6db5c802 Don't bother checking for new events from a source if the stream token hasn't advanced for that source 2015-05-13 15:08:24 +01:00
Mark Haines
63878c0379 Don't bother checking for updates if the stream token hasn't advanced for a user 2015-05-13 13:42:21 +01:00
Mark Haines
cffe6057fb Merge branch 'notifier_unify' into notifier_performance
Conflicts:
	synapse/notifier.py
2015-05-12 16:37:50 +01:00
Mark Haines
4429e4bf24 Merge branch 'develop' into notifier_unify
Conflicts:
	synapse/notifier.py
2015-05-12 15:31:26 +01:00
Mark Haines
78672a9fd5 Merge branch 'notifier_unify' into notifier_performance 2015-05-12 13:11:54 +01:00
Mark Haines
2551b6645d Update the end_token correctly, otherwise the token doesn't advance and the client gets duplicate events 2015-05-12 11:54:18 +01:00
Mark Haines
5c75adff95 Add a NotifierUserStream to hold all the notification listeners for a user 2015-05-12 11:00:37 +01:00
Mark Haines
e269c511f6 Don't bother passing the events to the notifier since it isn't using them 2015-05-11 15:01:51 +01:00
Mark Haines
5e3b254dc8 Use wait_for_events to implement 'get_events' 2015-05-11 14:37:33 +01:00
Erik Johnston
4ac1941592 PEP8 2015-05-08 16:33:01 +01:00
Erik Johnston
476899295f Change the way we do logging contexts so that they survive divergences 2015-05-08 16:32:18 +01:00
Erik Johnston
42c12c04f6 Remove some run_on_reactors 2015-05-01 14:41:25 +01:00
Erik Johnston
e19f794fee Change from exception to warn 2015-04-15 15:12:57 +01:00
Mark Haines
1280a47fc6 Add comment 2015-04-09 11:42:21 +01:00
Mark Haines
23d285ad57 Unset the timer in the timeout callback so that we don't try to cancel it if it has been called 2015-04-09 11:41:50 +01:00
Mark Haines
6f9dea7483 SYN-339: Cancel the notifier timeout when the notifier fires 2015-04-09 11:07:20 +01:00
Erik Johnston
5bc41fe9f8 Move comment into docstring 2015-04-08 14:01:22 +01:00
Erik Johnston
638be5a6b9 Factor out loops into '_discard_if_notified' 2015-04-08 13:58:32 +01:00